Surgical margins in breast conserving surgery for ductal carcinoma in situ of the breast and clinical outcomes: a national audit with long term follow-up.

Background: Optimal surgical margin width in breast conserving surgery (BCS) for ductal carcinoma in situ (DCIS) is not established. The United Kingdom (UK) Association of Breast Surgery (ABS) recommended a 1 mm margin, whereas a minimum of 2 mm has been recommended in the United States of America (USA). This paper uses precise histological margin width data from UK national datasets to understand the impact of surgical margins on time to recurrence (TTR). Methods: Patients were included if aged ≥45-years with a new diagnosis of DCIS alone, between 2003 and 2014, within the English National Health Service (NHS) Breast Screening Programme. Primary treatment included BCS and a minimum histological excision margin width recorded. Exclusion criteria included: prior history of DCIS; prior history of invasive cancer or its diagnosis within 3-months of initial surgical treatment for DCIS. Data was extracted from NHS England National Disease Registration Service (NDRS), ABS and Sloane Project audits. Findings: 16,907 patients diagnosed with DCIS having definitive BCS surgery were identified between 2003 and 2014. TTR was found to be significantly shorter for patients with surgical margins <1 mm vs ≥ 1 mm (adjusted hazard ratio (aHR) = 1·32; 95% (confidence interval) CI:1·06-1·63; p = 0·012); <2 mm vs ≥ 2 mm (aHR = 1·19; 95% CI:1·05-1·35; p = 0·0062) and ≥1-<2 mm vs ≥ 2 mm (aHR = 1·18; 95% CI:1·01-1·38; p = 0·032). There was no evidence that increasing the surgical margin width beyond 2 mm significantly improved TTR (aHR = 0·96; 95% CI: 0·86-1·08; p = 0·52 for ≥5 mm vs ≥ 2-<5 mm). The rate of recurrence across 14 years following BCS + radiotherapy was 1·2% per annum, 2129 (13%) patients had a recurrence of which 78% were invasive breast cancers. Interpretation: Patients with DCIS with histological margins of <2 mm, adjusted for other clinical factors, have significantly worse TTR compared to margins ≥2 mm. These findings may inform optimum treatment of patients with DCIS. Funding: An ABS grant covered the cost of data extraction by NHS England and medical writing assistance. The latter was provided by Edge Health, supervised by the co-authors.